ML20212M584
| ML20212M584 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 03/09/1987 |
| From: | Wilson R GENERAL PUBLIC UTILITIES CORP. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| RTR-REGGD-01.121, RTR-REGGD-1.121 5000-87-1181, 5211-87-2055, NUDOCS 8703120085 | |
| Download: ML20212M584 (10) | |
Text
e (O -O M GPU Nuclear NggIgf 100 Interpace Parkway Parsippany. New Jersey 07054 201 263-6500 TELEX 136-482 Writer's Direct Dial Number:
March 9, 1987 5000-87-1181 5211-87-2055 U.S. Nuclear Regulatory Commission Attention: Document Control Desk Washington, DC 20555 Gentlemen:
Three Mile Island Nuclear Generating Station (TMI-1)
NRC Denial of TSCR 148 Based on Use of Fracture Mechanics Methods to Justify 0TSG Tube Plugging Criteria In your letter of December 23, 1986 you informed GPU Nuclear Corporation that the NRC Staff had concerns about the fracture mechanics methodology used by GPUN to justify a 70% Once Through Steam Generator (OTSG) tube repair limit, and that Technical Specification Change Request (TSCR) 148 was therefore denied.
We maintain that the analytical basis which we provided in support of TSCR 148 is appropriate and correct, as detailed in the attached discussion.
We take exception to the Staff's conclusion that the GPUN_ analyses "contain significant errors and questionable assumptions". The assumptions used are reasonable on the bases of state of the art methodology subsequent to the issuance of Regulatory Guide 1.121.
We do not believe, however, that a public hearing is the appropriate forum for resolution of these types of technical differences with the NRC Staff. For this reason we are not requesting a hearing on the Denial of TSCR 148. GPUN has initiated efforts with the Electric Power Research Institute and various committees of the Babcock and Wilcox Owners Group with respect to application of Regulatory Guide 1.121 to OTSGs and the development of repair limits for degraded steam generator tubes. We are optimistic that further discussions with the NRC Staff will provide a basis for submittal of a similar Technical Specification Change Request in the future, after the completion of Owners Group efforts.
h 0703120005 070307 PDR ADOCK 05000209 p
PDR g
g i
GPU Nuclear is a part of the General Pubhc Utikties System
2-In the meantime, GPUN is preparing a separate and different Technical Specification Change Request for a permanent 50% throughwall repair limit, for defects of limited circumferential extent, which will be provided on a schedule to support TMI-1 Outage 7R (approximately July 1988). The basis for that Technical Specification Change Request is not expected to significantly depend upon those aspects of the structural mechanics i
methodology with which Staff has expressed reservations.
ry ruly yours, F.
Ison Vice President Technical Functions RFW/SK/pa(18619) cc:
J. F. Stolz
Attachment:
GPUN Discussion - NRC Denial of TSCR 148
GPUN DISCUSSION - NRC DENIAL OF TSCR 148 The Nucicar Regulatory Commission has denied a request by GPU Nuclear Corporation for an amendment to the Three Mile Island Unit 1 Operating License which would revise the provisions in the Technical Specifications relating to the steam generator tube plugging limitations in accordance with the licensees' application for amendment (TSCR 148) dated November 6, 1985.
Basically, the present Technical Specifications require repairing or removing from service a steam generator tube when a defect exceeds 40% of the tube wall thickness. The proposed amendment would maintain the 40%
through-wall limit on the secondary side of the tube but replace the limit on the primary side of the tube with a sliding scale which goes from 40%
to 70% through-wall for sufficiently short length defects.
The NRC has conducted an evaluation of the fracture mechanics and net section collapse methodology used by GPUN to justify a revised tube plugging limit. The NRC has, in their letter denying TSCR 148 dated December 23, 1986 and in a Safety Evaluation (SE) attached to the Denial, concluded that the GPUN analyses are not technically acceptabic. More specifically, the Denial states that the NRC Staff concludes that the fracture mechanics analyses contain significant errors and questionable assumptions in the areas of load development, material properties and analytical conclusions. The NRC Staff's positions on certain aspects of the fracture mechanics analyses are provided in the SE.
The purpose of this discussion is to address and comment upon the conclusions and positions expressed in the letter denying TSCR 148 and in the attached SE.
While the Staff has concluded that the GPUN analyses contain significant errors, it is GPUN's position that to the best of our knowledge the analyses as submitted in support of the amendment do not contain significant errors. Further, GPUN believes that the assumptions, methods, and material properties used along with the conclusions drawn from these analyses are reasonabic and justified.
Overview The proposed amendment, TSCR 148, was transmitted to the NRC on November 6, 1985. Two documents were submitted as part of TSCR 148 as the technical basis for the structural adequacy of the proposed tube plugging criteria. GPUN report TDR-645, " Bases for Plugging and Stabilizing Criteria for OTSG Tubes" was first submitted to the Commission by GPUN Letter 5211-85-2023 dated January 31, 1985, and a copy was provided with TSCR 148 as Appendix A.
A second GPUN report, TDR-690, " Comparison of GPUN Proposed OTSG Tube Plugging Criteria to Regulatory Guide 1.121" was submitted with TSCR 148 as Appendix B.
-2.
As pointed out in the TSCR 148 submittal, the proposed tube plugging criteria were based on analyses described in TDR-645. Further, the
' proposed plugging criteria were developed frou-several.~ existing _ analyses of the serviceability of flawed tubes under normal, transient or accident-conditions. These analyses included ASME Section III and Section XI:
evaluations, and a solid mechanics single accident load (Main Steam Line Break Accident, MSLB) analysis conducted as part of GPUN's response to the 1981 tube cracking experience (Ref. 1). The analyses described in TDR-645 show that the proposed tube plugging criteria would assure that small defects left in service would not grow through wall and that their behavior would be bounded by the Section III fatigue assessment, the Section XI crack growth evaluation and the MSLB solid mechanics analysis.
The stated objective of the second GPUN report, TDR-690, was to evaluate the proposed plugging criteria presented in TDR-645 against the guidelines of USNRC Regulatory Guide 1.121. Regulatory Guide 1.121 recommends a-margin of safety under normal operating and faulted conditions.
l-Regulatory Guide 1.121 does not provide specific guidance as to how recommended margins are to be evaluated with respect to solid mechanics.
GPUN chose to use state-of-the-art solid mechanics methods developed L
subsequent to issuance of Regulatory Guide 1.121 to evaluate recommended margins. The solid mechanics methods used were based on limit load or net l
section collapse theory supported with tube burst test data available at i
that time. Tearing instability fracture mechanics analyses were also performed as an' additional confirmation that net section collapse and not tearing instability would control tube failure.
The NRC Staff's Safety Evaluation of TSCR 148 is only related to GPUN report TDR-690 and Appendices to that report. The analyses described in TDR-645, which formed the bases for the proposed tube plugging criteria, do not appear to have been addressed by the SE.
In the following sections, comments and conclusions of the SE relative to the analyses ~ performed by GPUN in support of proposed tube plugging j
criteria as described in TSCR 148 are addressed.
Obj,ectiveofGPUNLetterofOctober3,1986 In the background section of the SE, reference is made to the GPUN letter of October 3, 1986 (Ref. 2). Referring to this letter, the SE states that "The licensee concluded that because of design and performance differences of the B6W OTSG, Regulatory Guide 1.121 methodology for determining i
allowable minimum wall thickness in stear. generator tubes does not apply."
i
{
I r
i l
m h1
-3, The purpose of the GPUN' letter of October 3, 1986, was to point out; differences between once-through steam generators, such as those at NI-1,
'and recirculation steam generators used elsewhere, and to provide.
additional insight as to how GPUN evaluated the margins recommended;in Regulatory Guide 1.121.- The GPUN letter concludes that, ".... Regulatory Guide 1.121 is not directly applicable in' casas where significant secondary stress exist. We believe a proper approach is to put margin on total' strain." The technical reasons for applying margins to strains rather than to loads (or stresses) are described in the next section.
GPUN agrees with the staff's statement:that " basic engineering guidance contained in Regulatory Guide 1.121 is applicable to all steam generators regardless of individual design and performance-induced difference" and
-GPUN has used this guidance to evaluate proposed tube plugging criteria.
As described in GPUN Report TDR-690, the Regulatory Guide 1.121 guidelines, including margins against structural failure, were used by GPUN in their evaluations.
i Load Development Source of Tube Loads As stated in the SE, GPUN used axial tube loads under normal operating and faulted conditions obtained from B4W topical reports as input for the structural analyses.- These reports, B4W-1588 and B4W-10146 (Refs. 3 and
- 4) address the minimum required tube wall thickness for 177-FA once-through steam generators and were prepared in support of the 40% of wall tube plugging criteria for B6W designed plants. These reports were previously referenced by GPUN as part of the justification for the return to service of the WI-1 steam generators following the kinetic expansion repair process.
(Ref. 1)
Independent analyses performed by GPUN as part of the WI-1 steam generator recovery program confirm that the B4W tube loads are appropriate (Ref. 1).
Further, evaluation of margins associated with proposed tube plugging criteria were performed for a peripheral tube which would experience the largest axial load.
It sb uld be noted that these evaluations were performed to justify the return to service of the entire lengths of the 31-1 steam generator tubes. The evaluations do not pertain solely to the kinetically expanded zone.
In Supplement 1 to NUREG 1019, " Steam Generator Repair and Return to Operation", the NRC relied on the review of NRC consultants at Brookhaven National Laboratories. Attachment 7 to NUREG 1019 Supplement I which is entitled "Three Mile Island Unit 1 Steam Generator - LEFM and Load Characterization", describes the purpose of the Brookhaven review:
p e,
"To support the contention that these tubes were..
" serviceable, a detailed linear elastic fracture mechanics evaluation (LEFM) was performed to demonstrate that defects of a permissible; size are not. susceptible to unstable-
- fallure. '..The purpose of the - review was.to determine :if '
-the proper loadings were considered in the LEFM studies and' Lif these studies support the GPUN contention that cracks of a permissible size will not lead to unstable. failure."
. Attachment 7 concluded that the analyses were found to be acceptable and supporting the GPUN contentions concerning crack propagation.
Nature of Tube Loads The-major portion of once-through steam generator tube axial loads is due to displacement of the tubes caused by temperature differences between.the steam generator. tubes and shell during a cooldown transient. Tube displacement produces what is called a secondary stress by Section III of the ASME Code. Secondary stresses, by their nature, cannot exceed the tube material flow properties and tube loads associated with such stresses are correspondingly limited. Failure will not occur unless the ductility of the material is exceeded.
The SE correctly points out that GPUN made use of these facts when
~
establishing tube loads for use in the net section collapse solid mechanics analyses.. On the other hand, GPUN did not calculate " actual tube" loads as suggested by the SE but calculated tube loads. associated with 3.0 and 1.428 times the predicted tube strain associated with normal operation and faulted transients, respectively. These loads are larger than the actual tube loads for both events.
The SE states that the licensee is correct in the assumption that nonlinear calculated tube loads are lower in magnitude than associated pseudo-elastic values, but concludes that it is technically unjustified to input these loads into net section collapse analyses. As background on this issue it should be understood that net section collapse criteria have been used as a basis for ASME Code,Section XI, flaw evaluation criteria in nuclear plant piping. Here, it is explicitly stated that because plastic collapse is the anticipated failure mechanism, secondary stress is assumed to be relaxed at failure and only primary stresses are used when performing flaw evaluations in accordance with Section XI, paragraph IWB-3610 (Ref. 5). A recent assessment of steam generator tube leak-before-break analysis methods also points out that in applying a limit-load based failure model (net section collapse), stresses derived from purely displacement controlled stress states do not contribute to plastic collapse provided that these stresses are in equilibrium within the section under load (Ref. 6).
In spite of this guidance and precedence, GPUN pointed out in its letter of October 3, 1986 (Ref. 2) that it believed a proper approach is not to neglect the secondary stresses, but to put a margin on total strain and demonstrate that the component has the capacity to resist the resulting stress and not fail.
~
- p. :
Flow-Induced Vibration Loads f
~
In the SE, the Staff has asserted that bending loads'due to flow induced vibration during a MSLB should have been addressed. As discussed during
.GPUN's presentation to the NRC on May 15, 1986, GPUN~ determined that the-A
' contribution of bending loads due to flow induced vibration is not
= controlling. -Axial loads are controlling, and do not occur concurrently.
.with flow induced vibration.. For these reasons, GPUN has concluded that M
flow induced vibration is not a technical issue.
Conclusion Considering tube load development, GPUN concludes that while it may be true that the axial tube loads (from Refs. 3'and 4) have not been verified by the NRC, tube loads used in the evaluation of proposed tube plugging criteria are reasonably bounded by existing analyses. Further, tube axial loads associated with 3.0 and 1.428 times the strain associated with the displacement produced by normal operating and faulted transients are justified and conservative. Using these loads in a net section collapse analysis is conservative in light of the guidance and precedence to neglect secondary stresses associated with displacement controlled loads.
Material Properties I
Tearing Instability Analyses As discussed previously, tearing instability analyses (J-T analyses) were carried out as an additional confirmation that net section collapse methods were suitable for evaluating the integrity of cracked Inconel 600 steam generator tubes. Margins associated with TSCR 148 proposed tube plugging 4
criteria were not directly evaluated using tearing instability methods and the unavailability of materials data needed for such analyses was clearly pointed out and discussed. We recognize that the use of the power law fit as representative of actual material behavior is in fact a representative and not necessarily conservative application. However, our purpose in j
presenting the tearing instability analyses was merely to demonstrate that net section collapse is clearly the controlling theoretical basis.
The assumption that net section collapse controls steam generator tube integrity has been made by other investigators (Ref. 6). This conclusion has also recently been confirmed with laboratory tests conducted by GPUN, results of which have not been provided to NRC. These laboratory tests were performed with 'IMI-1 archive tube material heat treated to be representative of in generator conditions. Circumferential1y oriented defects of different i
depth and length were machined from the tube inside surface and the tubes i
were axially loaded to failure. Failure loads confirm that net section collapse methods describe tube rupture.
i i
l
~
' " - Material' Flow Stress The SE correctly points out that material flow stress is a critical variable in net section collapse analysis. Here.GPUN used a. flow stress of 3Sm as
.recommen ed d in the literature (Ref. 5) and in Section XI'of the ASME Code.
For Inconel 600 material at operating temperatures this. stress is 69.9 ksi.
The SE asserts that an estimate of flow stress is inappropriate when. proven material yield and ultimate strength data are available. Using the data that GPUN provided from one test of one heat of material, the SE suggests that a flow stress of 63.5 ksi is appropriate. Further, the SE cites a reference which states that use of 3Se may not be a conservative estimate of flow stress for axially oriented flaws. And finally,.the SE concludes that in light of the approximately 10% difference between flow stress used by GPUN and assumed _by the Staff, the safety factors claimed (by GPUN) to be inherent in the proposed tube plugging criteria do not exist.
Use of net section collapse to demonstrate margins recommended by Regulatory Guide 1.121 does not require or even imply that lower bound material properties must be used. Rather, GPUN believes material data and methods should be reasonably representative. Additional margin on materials properties is not justified. Recently, published discussion of leak-before-break of steam generator tubes (Ref. 6) used an Inconel 600 material flow stress value of 80 ksi, 14% higher than the value used by
.GPUN. The axial crack burst test data referenced in the TSCR 148 submittal and the circumferential crack pull test data recently generated by GPUN with TMI-1 tube material both indicate that critical load predictions based on a 3Sm (as given in the ASME Code) flow stress are conservative.
An additional margin is contained in the GPUN net section collapse calculations for circumferential defects which the SE does not recognize.
That is, loads larger than those calculated to correspond to 3 times the normal operating and 1.428 times the faulted tube strains were used.
4 A
,-...n an.
,,,-n_-me----, - -, - - -, - -
,_,.,-.,,-,---c-.-n--c_,-
,,,,-n--e.,-n, e---.-
e,-
x ' Specifically, the load which has been predicted to cause failure of a tube with a 40% through-wall, full 360* of circumference, flaw (the existing tube plugging limit) was used in the calculations.
In conclusion, GPUN believes that the tearing instability analyses clearly i
confirm that net section collapse is the appropriate failure criterion.
GPUN believes that use of 3S,for the flow stress is reasonable and justified in net section collapse analyses.
Analytical Results for Axial Defects As pointed out in the TSCR 148 submittal, the analyses as presented in
. TDR-690 show that for relatively long axially oriented defects with greater than 60% through-wall penetration, margins as large as recommended by Regulatory Guide 1.121 are not met. GPUN pointed out that in their opinion this small discrepancy between analytical predictions and recommended margins is justified in light of the generally conservative nature of the calculations and favorable comparison between prediction and available test results.
The SE asserts that the test data cited in TDR-690 is too limited to support the conclusions that there is sufficient margin in calculations of axial defect behavior to account for the discrepancy noted above. The data was for Inconel 600 steam generator tubes with axial defects. The test data cited are for tubes with thickness equal or less than that of the TM1-1 steam generator tubes as well as greater. Also, the margins between test J
results and predictions are substantial (15 to 100%).
Summary
~
GPUN believes that solid mechanics analyses performed in support of proposed steam generator tube plugging criteria are reasonable and justified.
Further, GPUN contends that these analyses show that the intent and general guidance of Regulatory Guide 1.121 are satisfied.
i L
T 5
1 1
'. ' w 8-References n
1.
1R-008, " Assessment of 1MI-1 Plant Safety for Return to Service After Steam Generator Repair", Revision 3, September 1983 2.
GPUN Letter 5000-86-1050/5211-86-2174, " Regulatory Guide 1.121 Extent of Applicability to OTSG's", October 3, 1986-3.
B4W-1588 " Determination of Minimum Required Tube Wall Thickness for 177-FA OTSGs", April 1980 4.
B4W-10146 " Determination of Minimum Required Tube Wall Thickness for 177 FA OTSGs", October 1980.
5.
" Evaluation of Flaws in Austenitic Steel Piping", Journal of Pressure Vessel Technology, Volume 108, August 1986, pages 352-356.
6.
- Lang, J., F., et. al., " Analysis of Leak-Before Break for Steam Generator Tubes, Proceedings of Third International Topical Meeting on Reactor Thermal Hydraulics", Vol. 2, Sessions 13-21, C. Chong and G.
Brown, Eds, Newport, Rhode Island, October 15-18, 1985.
7.
Griesbach, T., et. al., Analysis Methods for Evaluating Leak-Before-Breaks in U-Tube Steam Generators, Ibid.
-